Hybrid Visual-Ranging Servoing for Positioning Based on Image and Measurement Features

Area of Science:

• Robotics
• Computer Vision
• Control Systems

Background:

• Traditional vision-based robotic control systems face challenges with local minima and singularity issues.
• Accurate high-precision positioning for robotic manipulators is crucial in various industrial applications.

Purpose of the Study:

• To propose a hybrid visual-ranging servoing method for high-precision positioning of a 6-degree of freedom (DOF) manipulator.
• To address and overcome limitations of traditional vision-based systems, such as local minima and singularity problems.

Main Methods:

• A hybrid visual-ranging servoing approach is developed, directly utilizing image and measurement features within the control loop.
• A full-rank interaction matrix hybrid visual servo (FRHVS) design criterion is introduced to ensure both the interaction matrix and its pseudoinverse are full rank.
• The interaction matrix, combining image and other sensor features, is derived analytically.

Main Results:

• The proposed method enables 6-DOF control of a robot without complex image feature design or attitude estimation.
• Experimental results on a 6-DOF manipulator demonstrate the method's effectiveness, global asymptotic stability, and high precision.

Conclusions:

• The hybrid visual-ranging servoing method offers a robust solution for high-precision robotic positioning tasks.
• The FRHVS design criterion successfully mitigates singularity and local minima issues inherent in vision-based control.